In recent years, server integration utilizing a virtualizing technique has been widely spreading. However, in the case where a virtual machine is transferred (migration) between servers having different performances in order to perform maintenance, for example, due to a high load or trouble after the server integration, it is difficult to accurately estimate the load of the virtual machine after the transferring because of difference in virtualization types, or difference in hardware performances between servers. As a method for solving this problem, various methods have been devised, in which the destination server simulates the load of the virtual machine to know the performances in advance.
As such a technique, Patent Document 1 (Japanese Patent Application Laid-open No. 2008-146313) discloses the following technique. Here, a description will be made with reference to FIG. 1 of Patent Document 1. An active server 10 is a computer that performs an actual job operation. A standby server 20 is a computer that takes over the actual job operation from the active server 10 in the case where any trouble occurs in the active server 10. A newly installed server 30 is a computer that is to be newly installed. A management terminal 40 is a computer used by a manager who evaluates performances of the newly installed server 30. A communication network 50 is, for example, the Internet or a local area network (LAN).
The management terminal 40 receives an input of a benchmark program that provides an input data generating function of generating input data and a CPU load generating function of using a CPU of the computer in accordance with the input data. Then, the standby server 20 and the newly installed server 30 are caused to run the benchmark program. The standby server 20 measures a processing load of the standby server 20 itself caused as a result of running the benchmark program. The newly installed server 30 measures a processing load of the newly installed server 30 itself caused as a result of running the benchmark program. The management terminal 40 outputs the processing loads of the standby server 20 and the processing load of the newly installed server 30, thereby to realize a load simulation, by which the performance evaluation of the computer can be easily performed.
With the configuration described above, of application programs for the actual job operation running in the active server 10, an application program (benchmark program) simulating an operation of an application having a higher processing load is installed in the newly installed server 30, and, a processing load (amount of resource consumption) of the newly installed server 30 is measured at the time when the newly installed server 30 causes the benchmark program to process the input data same as that processed in the active server 10. This enables the newly installed server 30 to reproduce a processing load equivalent to the processing load of the application program for the actual job operation.
Further, Patent Document 1 discloses that similar processing is possible in the case of a virtual machine in lieu of the application program.